RAG4 Gene Encodes a Glucose Sensor in Kluyveromyces lactis

Corresponding author: Micheline Wésolowski-Louvel, Unité Microbiologie et Génétique, ERS 2009 CNRS/UCBL/INSA, Université Claude Bernard, Bâtiment 405, 43, Boulevard du 11 Novembre 1918, 69622 Villeurbanne Cédex, France., louvel{at}univ-lyon1.fr (E-mail)

Communicating editor: M. JOHNSTON

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

The rag4 mutant of Kluyveromyces lactis was previously isolated as a fermentation-deficient mutant, in which transcription of the major glucose transporter gene RAG1 was affected. The wild-type RAG4 was cloned by complementation of the rag4 mutation and found to encode a protein homologous to Snf3 and Rgt2 of Saccharomyces cerevisiae. These two proteins are thought to be sensors of low and high concentrations of glucose, respectively. Rag4, like Snf3 and Rgt2, is predicted to have the transmembrane structure of sugar transporter family proteins as well as a long C-terminal cytoplasmic tail possessing a characteristic 25-amino-acid sequence. Rag4 may therefore be expected to have a glucose-sensing function. However, the rag4 mutation was fully complemented by one copy of either SNF3 or RGT2. Since K. lactis appears to have no other genes of the SNF3/RGT2 type, we suggest that Rag4 of K. lactis may have a dual function of signaling high and low concentrations of glucose. In rag4 mutants, glucose repression of several inducible enzymes is abolished.

FERMENTATION is the main mode of energy acquisition in Saccharomyces cerevisiae, and the redundancy of many glycolytic genes in this yeast may reflect its importance. By contrast, in Kluyveromyces lactis fermentation is dispensable, and glycolytic genes are not redundant. This situation allowed us to isolate many mutants defective in key genes of the fermentation pathway. The central role of glucose transporters in the regulation of fermentation has thus been demonstrated (CHEN et al. 1992 ; WESOLOWSKI-LOUVEL et al. 1992A ). A number of trans-acting elements involved in this regulation have also been identified.

The glucose uptake system in K. lactis relies on two, nonredundant, glucose transporters with a few exceptions found in a variant set of strains: a low-affinity carrier encoded by RAG1 (WESOLOWSKI-LOUVEL et al. 1992A ) and a high-affinity carrier encoded by HGT1 (BILLARD et al. 1996 ). The expression of RAG1 is glucose inducible (CHEN et al. 1992 ; WESOLOWSKI-LOUVEL et al. 1992A ), whereas HGT1 is constitutively expressed (BILLARD et al. 1996 ). The expression of RAG1 is necessary for fermentative growth on high concentrations of glucose, because rag1 mutants are unable to grow on 5% glucose when respiration is blocked by antimycin A [this is called Rag^- phenotype (resistance to antimycin A on glucose); GOFFRINI et al. 1989]. A series of mutants displaying the Rag^- phenotype have been isolated (WESOLOWSKI-LOUVEL et al. 1992B ) and at least three of them carry a mutation in one of the three genes that positively regulate the transcription of the RAG1 gene (CHEN et al. 1992 ): RAG5, which encodes the single hexokinase of K. lactis (PRIOR et al. 1993 ); RAG8, encoding a caseine kinase I (BLAISONNEAU et al. 1997 ); and RAG4, the subject of this study. RAG4 belongs to the SNF3/RGT2 family of genes that, in S. cerevisiae, are thought to form part of the glucose-sensing apparatus (OZCAN et al. 1996 , OZCAN et al. 1998 ).

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Yeast strains and growth conditions:
The strains used in this study are as follows: MW270-7B (MATa uraA1-1 leu2 metA1-1 Rag⁺; BILLARD et al. 1996 ); PM6-7A/VV32 (MATa uraA1-1 adeT-600 rag4-1 Rag^-; WESOLOWSKI-LOUVEL et al. 1992B ); JA6 (MATa ura3 trp1 ade1-600 adeT-600 Rag⁺; BREUNIG 1989 ); MW109-8C/FA49 (MAT lysA1-1 rag4-5 Rag^-; WESOLOWSKI-LOUVEL et al. 1992B ); JA6/29-4 [MATa ura3 trp1 ade1-600 adeT-600 kht1(=rag1) kht2 Rag^-]; JA6/151 (MATa ura3 trp1 ade1-600 adeT-600 rag5 Rag^-; GOFFRINI et al. 1995 ).

Yeast cells were grown at 28° either in a complete medium containing 1% Bacto yeast extract, 1% Bacto-peptone (Difco, Detroit), supplemented with glucose (as indicated) or in a minimal medium containing 0.7% yeast nitrogen base without amino acids (Difco) but with auxotrophic supplements as required and a specified carbon source. The Rag phenotype was tested on GAA medium (5% complete glucose medium supplemented with 5 µM antimycin A). The 2-deoxy-D-glucose- resistant phenotype Dgr⁺ was tested in a minimal medium containing 2-deoxy-D-glucose (Sigma, St. Louis) at the concentration of 5 g liter^-1. The different carbon sources utilized were added at 2% (w/v).

Genetics methods have been described previously (WESOLOWSKI et al. 1982 ; GOFFRINI et al. 1989 ).

Yeast transformation:
Replicative transformation of K. lactis was performed by electroporation. For integrative transformation of K. lactis, the procedure described by DOHMEN et al. 1991 was followed.

Cloning and sequencing of RAG4 gene:
The RAG4 gene was cloned by in vivo complementation of rag4-1 mutation (strain PM6-7A/VV32) with a K. lactis genomic library made in the KCp491 vector. Out of 13,000 Ura⁺ transformants, two were found to be Rag⁺. The complementing plasmids extracted from these two transformants were then amplified in Escherichia coli. In each case, amplified DNA showed an expected pattern (vector + insert) for only one preparation out of three. In the other cases, the restriction profiles of DNA suggested loss of the insert together with reorganization of the vector. Further large-scale preparation of the complementing DNA turned out to be impossible, whatever the E. coli host (MC1066, XL1-Blue, and Sure). The complementing DNA that we had obtained in small quantities allowed us to establish that the two Rag⁺ transformants carried the same plasmid pSB1 containing a 6.0-kb insert (Fig 1). To save this DNA preparation, part of the restriction map of pSB1 was obtained from PCR-amplified DNA. The oligonucleotides used were as follows: TetBup, TCCTGCTCGCTTCGCTACTTGG (upstream of the BamHI site of tetracycline marker in which was cloned the genomic DNA of K. lactis) and TetBlw, CCATACCCACGCCGAAACAAGC (downstream of the BamHI site). Then, subfragments of the amplified insert were cloned into pCXJ18 vector and tested for their stability in E. coli as well as for their ability to complement the rag4-1 mutation. The results (Fig 1A) indicated that the largest subfragment (4.4-kb PvuII-BamHI) stable in E. coli did not complement the rag4 mutation. Therefore, the 2.7-kb PvuII-EcoRI fragment of the pSB1 plasmid (original DNA) cloned into SmaI and EcoRI sites of pCXJ18 plasmid was entirely sequenced on both strands. The 5' end of the RAG4 gene (upstream of the PvuII site) was sequenced after PCR amplification, using the Expand Long Template PCR system (Boehringer, Mannheim, Germany) with pSB1 original DNA as template. The oligonucleotides used were as follows: TetBup (see above) and RG-41, TGACGTGAACGATGTTCA (starting 100 nucleotides downstream of the PvuII site of pSB1 DNA). The resulting 1.8-kb PCR product was sequenced on both strands.

View larger version (8K): In this window In a new window Download PPT slide   Figure 1. (A) Restriction map of recombinant plasmid pSB1 carrying the RAG4 gene and other subclones. , open reading frame; + and -, presence or absence of complementation of rag4 mutation. (B) Internal fragment disruption of the RAG4 gene with URA3 marker (shaded box). The open and solid boxes indicate the genomic fragment. Solid boxes represent the RAG4 sequence (see MATERIALS AND METHODS).

The RAG4 nucleotide sequence has been assigned the EMBL accession no. Y14849.

Construction of rag4 deletion strains:
The 2.7-kb PvuII-EcoRI fragment containing most of the 3' region of the RAG4 gene and cloned into the pCXJ18 plasmid was recloned into the pBluescript KS phagemid (Stratagene, La Jolla, CA), using the BamHI and the EcoRI sites (Fig 1B). The resulting plasmid was digested with StyI, blunt ended with Klenow enzyme prior to digestion with BglII. Then, the internal 1.5-kb StyI-BglII fragment was replaced by a 1.0-kb SphI-EcoRI fragment that contained the URA3 marker from pAF101 vector (provided by B. Dujon, Institut Pasteur, Paris; THIERRY et al. 1990 ). A 2.2-kb BamHI-EcoRI fragment that contained the disrupted RAG4 cassette was used to transform two Rag⁺ hosts MW270-7B and JA6 to uracil prototrophy. Correct integration of the disrupted gene was verified by Southern hybridization (data not shown).

Complementation of the rag4 mutation by SNF3 and RGT2:
The two genes were cloned in low-copy-number URA3-marked vectors of K. lactis. The 4.1-kb HindIII-NruI fragment of the SNF3-containing plasmid pBL8 (provided by M. Carlson, Columbia University, New York; MARSHALL-CARLSON et al. 1990 ) was cloned into the HindIII-NruI site of KCp491 (PRIOR et al. 1993 ). In the case of RGT2, a 3.3-kb EcoRI-BamHI fragment of plasmid pBM3272, provided by S. Özcan (Washington University, St. Louis), containing the entire RGT2 gene was cloned into the EcoRI-BamHI site of pCXJ18 (CHEN 1996 ). The constructs were individually transformed into the original rag4-1 strain (PM6-7A/VV32). The Ura⁺ transformants thus obtained were then tested for their Rag phenotype by replica-plating on 5% glucose + antimycin A (GAA) plates.

Preparation of yeast RNA and probes:
Total RNA was extracted from cells grown to an OD₆₀₀ of 2 to 3. Poly(A)⁺ enriched mRNA were obtained using an mRNA separator (CLONTECH, Palo Alto, CA). The RAG4 probe used was the 0.98-kb PvuII-KpnI fragment (Fig 1) of pSB1 DNA. The RAG1 probe used was a specific 0.9-kb SalI-PstI internal fragment of the gene (WESOLOWSKI-LOUVEL et al. 1992A ). This probe that was also used to analyze KHT1 (identical to RAG1) and KHT2 (a variant of RAG1) transcription in JA6 strain does not discriminate between the two transcripts. The HGT1 probe was a 1.75-kb EcoRI-HindIII fragment containing the HGT1 gene (BILLARD et al. 1996 ). KlGAL80 and LAC9 probes were obtained by PCR amplification using K. lactis genomic DNA as template. The oligonucleotides used were as follows: LAC9a, ATGGGTAGTAGGGCCTCCAATTCG; LAC9b, CACTGTTCGTACCAGTGGTCTC; GAL80a, CGGCAGGACGGCATCATCATGAAC; and GAL80b, GAGGACATGGCAACATTAAG. The LAC4 probe used was the plasmid pLX8 (provided by K. D. Breunig, Martin-Luther-University, Halle-Wittenberg, Germany) containing the whole LAC4 gene. KlCYB2 probe was a 1.9-kb SalI-EcoRI fragment from p30 plasmid (ALBERTI et al. 2000 ). In all cases a specific probe of K. lactis actin gene was used in parallel as a quantitative reference.

Measurement of glucose uptake:
Glucose uptake was measured as previously described (BILLARD et al. 1996 ). Uptake activity was determined by a 10-sec incubation with a ¹⁴C-labeled glucose at each glucose concentration.

Preparation of cell-free extract and enzyme assays:
Cells, harvested at a density of 5 x 10⁷ cells ml^-1, were resuspended in extraction buffer (0.1 M Tris-HCl, pH 8.5, 1 mM phenylmethylsulfonyl fluoride) and disrupted by vortexing at 4° in the presence of glass beads. The supernatant was utilized for measurement of the enzyme activity.

ß-Galactosidase (EC 3.2.1.23) activity was assayed as described by COHN and MONOD 1951 , and L-lactate ferricytochrome c-oxidoreductase (L-LCR; EC 1.1.2.3.) according to LODI et al. 1994 . Total -glucosidase (maltase, EC 3.2.1.20 and isomaltase, EC 3.2.1.10) activity was tested with p-nitrophenyl-O-D-glucopyranoside as a substrate as described by ZIMMERMANN et al. 1977 . The Lowry method was used for protein quantification with bovine serum albumin as standard.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Isolation of the RAG4 gene and deduced amino acid sequence of its product:
We isolated a plasmid (pSB1) from a CEN-based K. lactis genomic library by its ability to complement the rag4-1 mutation and confer the Rag⁺ phenotype (see MATERIALS AND METHODS). The nucleotide sequence of the left end of the 4.4-kb PvuII-BamHI fragment revealed the presence of an open reading frame (ORF) whose putative product showed a high homology with SNF3 and RGT2 gene products of S. cerevisiae. This ORF should correspond to the RAG4 gene since the right half of the insert was not complementing. The complete nucleotide sequence of the RAG4 gene was established as reported in MATERIALS AND METHODS. The deduced Rag4 protein was found to be 716 amino acids long, of which the first 113 were localized to the left of the PvuII site of pSB1 DNA (Fig 1). Probably it is this or flanking region of the RAG4 gene that, for unknown reasons, contained the element(s) of instability (see MATERIALS AND METHODS). The predicted Rag4 protein presents 12 potential transmembrane domains, typical of sugar permeases, as well as a 251-residue C-terminal tail. This C-terminal extension is characteristic of the glucose sensor proteins of S. cerevisiae Snf3 and Rgt2, which have been shown to control the transcription of several glucose permease genes (LIANG and GABER 1996 ; OZCAN et al. 1996 ). Rag4 has the same level of similarity (72–74%) and identity (52–53%) with both proteins. The highest level of identity between these three proteins is localized within the transmembrane regions (63–66%). In the three proteins, the carboxy-terminal extensions are rather dissimilar, except for a sequence motif of 25 amino acids that occurs twice in Snf3 and only once in Rgt2 and Rag4. This sequence has been shown to be essential for the signaling function in S. cerevisiae (OZCAN et al. 1998 ; VAGNOLI et al. 1998 ).

Disruption and expression of RAG4 gene:
A rag4 null mutant is Rag^-. Moreover, the allelism of the disrupted gene with rag4 was confirmed by the absence of complementation in a cross between the disruptant and the rag4-5 strain (MW109-8C/FA49). Thus, the cloned gene indeed corresponds to the RAG4 locus, and the rag4 null mutation is not lethal.

Transcript level of RAG4 was found to be very low, and the mRNA could be detected only when the poly (A)⁺ fraction of the total RNA was used (Fig 2). SNF3 and RGT2 genes of S. cerevisiae are also known to be transcribed at low level (NEIGEBORN et al. 1986 ; OZCAN et al. 1996 ). The RAG4 transcript is 2.8–2.9 kb long, consistent with the size of the ORF. The level of RAG4 mRNA did not change significantly with the carbon sources of the growth media (2% glycerol or 2% glucose; Fig 2).

View larger version (52K): In this window In a new window Download PPT slide   Figure 2. Northern blot analysis of RAG4 mRNA. Each slot was loaded with 5 µg of poly(A)+ RNA and electrophoresed on a 1.2% agarose-formaldehyde gel. The probes used are described in MATERIALS AND METHODS. Lane 1, MW270-7B/rag4 null mutant strain grown on 2% glucose; lane 2, MW270-7B (RAG4) grown on 2% glycerol; lane 3, MW270-7B (RAG4) grown on 2% glucose.

Complementation of the rag4 mutation by RGT2 and SNF3 genes of S. cerevisiae:
The functional similarity of Rag4 and Snf3/Rgt2 was suggested by the impairment of transcription of glucose permease genes in the corresponding mutants: RAG1 gene transcription is affected in rag4 mutant in K. lactis (CHEN et al. 1992 ), as is transcription of several HXT genes in snf3 and rgt2 mutants of S. cerevisiae (OZCAN et al. 1996 ). The similarity of the Rag4 protein to Snf3 and Rgt2, as well as the similar effect of rag4 and snf3 and rgt2 on expression of glucose transporter genes prompted us to try to complement the rag4 mutation with the SNF3 and RGT2 genes. Each of the two genes was able to restore growth of the rag4 strain on glucose + antimycin A (Fig 3). This was unexpected since SNF3 and RGT2 have distinct sensor functions for glucose.

View larger version (42K): In this window In a new window Download PPT slide   Figure 3. Complementation of the rag4 mutation by RGT2 and SNF3 genes of S. cerevisiae. rag4 cells (strain PM6-7A/VV32) transformed with the centromeric vector pCXJ18 (control) or with the SNF3 and RGT2 genes carried by KCp491 and pCXJ18, respectively. The transformants were grown on 2% glucose uracil-less medium prior to being replica plated onto GAA medium. The growth on GAA plates (Rag+/Rag- phenotype) was checked after 24 hr of incubation at 28°. RAG4 strain (MW270-7B) was used as a Rag+ control.

Growth phenotype and glucose uptake in the rag4 null mutant:
The S. cerevisiae snf3 mutant but not rgt2 is defective in growth on raffinose (low glucose) plus antimycin (NEIGEBORN et al. 1986 ; SCHMIDT et al. 1999 ). The snf3 rgt2 double mutant does not grow on raffinose or glucose (0.2 and 2%) plates containing antimycin A and grows poorly on 2% glucose even in the absence of antimycin A (OZCAN et al. 1998 ; SCHMIDT et al. 1999 ). We examined in more detail the growth of the rag4 null mutant on low glucose (0.1%) and high (2%) glucose, as compared to its isogenic wild-type strain. Fig 4 shows that the growth of the rag4 null mutant was reduced at both concentrations of glucose. The phenotype of the rag4 mutant on glucose thus appears to be similar to that of the snf3 rgt2 double mutant of S. cerevisiae.

View larger version (56K): In this window In a new window Download PPT slide   Figure 4. Growth phenotype of the rag4 null mutant strain. The two strains MW270-7B (RAG4) and MW270-7B/rag4 (rag4::URA3) were streaked to single colonies on 0.1 and 2% glucose minimal plates and incubated for 3 days at 28° before the photographs were taken.

In the rag4 mutant, the glucose-induced expression of the low-affinity glucose permease gene RAG1 is severely reduced (CHEN et al. 1992 ; this study). We therefore measured the actual glucose uptake activities in the rag4 null mutant. As shown in Fig 5, both low- and high-affinity glucose transport activity is reduced in the mutant, although not completely absent. Therefore, the reduced growth rate on glucose of the rag4 mutant may be explained by its low glucose uptake activity.

View larger version (10K): In this window In a new window Download PPT slide   Figure 5. Glucose uptake in rag4 null mutant. At various concentrations of glucose (0.5–80 mM), the rate of D-[14C]glucose uptake was determined by a 10-sec incubation. Uptake velocity, V, expressed as nanomoles of glucose per milligram (dry cell mass) per minute, was plotted against V/S (S is the millimolar glucose concentration).

The expression of glucose-permease genes is controlled by Rag4:
As in the original rag4 strain, the transcription of RAG1 could not be detected in the null mutant (rag4), while that of HGT1 was slightly increased in the mutant (Fig 6A and Fig B). Most K. lactis strains harbor the two single genes RAG1 and HGT1 coding for low- and high-affinity glucose permeases, respectively. However, some natural isolates do not contain the HGT1 gene (our unpublished data), and their RAG1 locus is replaced by two tandemly arranged glucose transporter genes, KHT1 (identical to RAG1), inducible by high levels of glucose, and KHT2 (a variant of RAG1), weakly induced by low glucose and repressed by high glucose (WEIRICH et al. 1997 ; BREUNIG et al. 2000 ). In this type of strain, the transcripts of both genes are practically absent in rag4 mutant cells grown either at low or high glucose concentration (Fig 6C). Therefore, Rag4 appears to be required for the specific glucose-induced expression of both of these genes. It is noteworthy that, in this case, induction by both low and high concentrations of glucose is dependent on Rag4.

View larger version (21K): In this window In a new window Download PPT slide   Figure 6. Effect of the disruption of RAG4 on the transcription of glucose-permease genes. Approximately 15 or 25 µg of total RNA extracted from cells grown on glucose complete medium was loaded in each slot. Electrophoresis conditions were as in Fig 2. The probes used are described in MATERIALS AND METHODS. (A) RAG1 transcription. The cells were grown on 2% glucose medium. MW270-7B (RAG4) strain (lane 1); isogenic MW270-7B/rag4 strain (lane 2). (B) HGT1 transcription. The cells were grown on 2% glucose medium. MW270-7B (RAG4) strain (lane 1); isogenic MW270-7B/rag4 strain (lane 2). (C) KHT1 and KHT2 transcription. Total cellular mRNA was prepared from cells grown on 0.2% (lanes 1 and 2) and 2% glucose (lanes 3 and 4). JA6 (RAG4) strain (lanes 1 and 3); isogenic JA6/rag4 strain (lanes 2 and 4).

Role of RAG4 gene in glucose repression:
The K. lactis strains most sensitive to glucose repression are those mentioned above that contain the two glucose transporter genes KHT1 and KHT2 (GOFFRINI et al. 1995 ; BREUNIG et al. 2000 ). Therefore, we used the JA6 wild-type strain and its isogenic derivatives to investigate the possible role of Rag4 in glucose sensing and glucose repression. To measure glucose repression, we tested the ability of 2-deoxy-D-glucose (2-DOG), a nonmetabolizable glucose analogue known to cause glucose repression, to inhibit growth on galactose, maltose (-glucosyl--glucose), raffinose (-galactosyl--glucosyl-ß-fructose), ethanol, and lactate (ZIMMERMANN and SCHEEL 1977 ; GOFFRINI et al. 1995 ). As shown in Fig 7, the rag4 mutant, but not wild type, was able to grow in the presence of the 2-DOG on lactose, galactose, lactate, and ethanol. Fig 7 also shows that both a kht1 kht2 mutant and a rag5 (hexokinase) mutant grew on all these carbon sources in the presence of 2-DOG. These results suggest that rag4 mutation causes a release from glucose repression of the enzymes involved in the utilization of lactose, galactose, lactate, and ethanol, but not the enzymes of maltose or raffinose metabolism.

View larger version (75K): In this window In a new window Download PPT slide   Figure 7. Effect of different rag mutations on deoxyglucose resistance phenotype. The strains were streaked on minimal media supplemented with 2% of the different carbon sources and 0.5% 2-deoxy-D-glucose. The plates were incubated for 3 days at 28° before the photographs were taken.

The levels of the enzymes necessary for the utilization of lactose/galactose and L-lactate [ß-galactosidase and L-lactate ferricytochrome c-oxidoreductase (L-LCR), respectively] were completely derepressed in cells grown on glucose (Fig 8). Maltase activity was derepressed in the kht1 kht2 mutant, but not in the rag4 mutant. These results are fully consistent with the growth phenotypes observed in the presence of 2-DOG.

View larger version (51K): In this window In a new window Download PPT slide   Figure 8. Glucose repression in JA6/rag4 and JA6/29-4 (kht1 kht2) mutants. Cultures were grown in YP medium containing the specific inducers galactose (ß-galactosidase) and lactate (L-LCR and maltase), with or without the addition of 2% glucose. (It has been previously demonstrated that in JA6 strain the maximal induction of maltase is observed in lactate; GOFFRINI et al. 1995 .) (A) ß-Galactosidase activity; (B) L-LCR activity; (C) maltase activity. Enzyme activities were measured as described in MATERIALS AND METHODS. Enzyme activity units (EAU) are expressed as nanomoles of substrate per minute (per milligram of protein). All values are means of three independent experiments. In no case was the variation >15%.

The derepression phenotypes observed in rag4 are mediated by changes in the transcription of the corresponding genes, since the levels of LAC4 and KlCYB2 mRNA are strongly reduced by glucose in wild type, but not in the rag4 mutant (Fig 9A).

View larger version (52K): In this window In a new window Download PPT slide   Figure 9. Northern blot analysis of LAC/GAL and KlCYB2 genes in rag4 null mutant. Total cellular mRNA was prepared from cells grown on 2% galactose (lanes 1 and 3) and 2% galactose plus 2% glucose (lanes 2 and 4). JA6 (RAG4) strain (lanes 1 and 2); isogenic JA6/rag4 strain (lanes 3 and 4). (A) LAC4 transcription. (B) KlCYB2 transcription. (C) KlGAL80 transcription. (D) LAC9 transcription. The probes used are described in MATERIALS AND METHODS.

The effect of Rag4 on glucose repression of LAC4 is probably mediated through one of the two regulatory genes of the lactose-galactose regulon: LAC9 (KlGAL4), the transcriptional activator (SALMERON and JOHNSTON 1986 ), and LAC10 (KlGAL80), the inhibitor of KlGAL4 (DICKSON et al. 1981 ; ZENKE et al. 1993 ). Rag4-dependent repression of the lactose-galactose regulon could result either from a positive effect on the expression of LAC10 or from a negative effect on the activator gene LAC9. The expression of KlGAL80 was not affected by rag4 mutation both in repressing (glucose + galactose) and nonrepressing (galactose) conditions (Fig 9B). In galactose medium, LAC9 gene was transcribed to a similar level in the wild-type and rag4 strains. When glucose was present in the growth medium, LAC9 transcription was reduced in the parental strain (RAG4) but not in the rag4 mutant. We conclude that the glucose repression of LAC9 requires Rag4.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

The RAG4 gene has been supposed to have some regulatory function in the glucose uptake system in K. lactis because, in this mutant, the induced transcription of the glucose transporter gene RAG1 is lost (CHEN et al. 1992 ; WESOLOWSKI-LOUVEL et al. 1992B ). The transcription of the high-affinity glucose permease gene HGT1 is also modified in a rag4 mutant (BILLARD et al. 1996 ). RAG4 codes for a typical protein of the transmembrane sugar transporter family and shares a high level of identity with two S. cerevisiae proteins, Snf3 and Rgt2. Both of these proteins have been proposed to act as glucose sensors that control hexose transporter gene expression in this yeast (OZCAN et al. 1996 , OZCAN et al. 1998 ). Like these proteins, Rag4 harbors a characteristic long C-terminal extension that has been shown to be important for glucose signaling function by Snf3 and Rgt2 (COONS et al. 1997 ; OZCAN et al. 1998 ; VAGNOLI et al. 1998 ). Also, like the two S. cerevisiae genes, RAG4 was found to be transcribed at a low level, which is consistent with a supposed regulatory function of the product rather than a transport activity.

Snf3 has been proposed to function as a sensor of low levels of glucose and Rgt2 as a sensor of high glucose levels (OZCAN et al. 1996 , OZCAN et al. 1998 ). As the RAG4 gene product regulates positively the low-affinity glucose transporter gene RAG1 (CHEN et al. 1992 ; this study), one may suppose a priori that Rag4 is a high glucose sensor like Rgt2 of S. cerevisiae. However, we found that either SNF3 or RGT2 can fully complement in vivo the rag4 mutation of K. lactis. Therefore, there is the possibility that Rag4 is performing both high and low glucose-sensing functions in K. lactis. However, the capacity of Snf3 to complement the rag4 mutation of K. lactis could be due to the ability of the high-affinity glucose sensor to sense high glucose concentrations. To verify whether Rag4 really has a double function, an obvious test is complementation of snf3 and rgt2 mutations by RAG4. It is unfortunate that the unusual instability of the RAG4 gene-carrying plasmids in E. coli hosts makes this experiment very difficult at this time. The element(s) responsible for this unexplained instability seems to reside within or close to the 5' end of the RAG4 gene. Nevertheless, several facts are in favor of the hypothesis that Rag4 has a dual sensing function. First, the growth ability of the rag4 mutant was altered at both low and high glucose, as it is in the snf3 rgt2 double mutant of S. cerevisiae (SCHMIDT et al. 1999 ). Second, high- and low-affinity glucose transport were both affected in a rag4 mutant. Third, the transcription of the low- and high-affinity glucose transporter genes RAG1 and HGT1, respectively, was impaired or increased in the rag4 strain of the CBS 2359 genetic background. In the JA6 genetic context where the single gene RAG1 has been replaced by KHT1 plus KHT2, the transcription of the two genes was no longer induced by high and low levels of glucose, respectively. Fourth, disruption of RAG4 in a glucose- responsive strain (JA6) abolished glucose repression of several genes (like LAC4 and KlCYB2), whereas in S. cerevisiae glucose repression of SUC2 and GAL1 can be prevented only in the snf3 rgt2 double mutant (OZCAN et al. 1998 ; SCHMIDT et al. 1999 ). Finally, we found no other SNF3/RGT2-related sequences in K. lactis despite our attempts to detect such gene(s) using a probe for the specific 25-codon sequence (the supposed glucose sensor signature) common to the carboxy tail of RAG4, SNF3, and RGT2 (Southern hybridization experiments at low stringency; data not shown). Therefore, all data so far available support the idea that K. lactis has a single glucose sensor, Rag4, which is able to detect low and high levels of glucose and mediate the signal to glucose transporter genes.

Although less pronounced than in S. cerevisiae, glucose repression is also an important regulatory device in some strains of K. lactis (JA6 series; FERRERO et al. 1978 ; BREUNIG 1989 ; GOFFRINI et al. 1995 ). When the two glucose permease genes KHT1 and KHT2 are deleted, glucose repression of lactose regulon enzymes as well as several other enzymes required for the utilization of lactate, ethanol, maltose, and raffinose (Fig 8) is abolished (WEIRICH et al. 1997 ). This confirms that glucose uptake is a crucial parameter for glucose repression. The finding that the rag4 mutation, despite the severe inhibition of KHT1 and KHT2 expression, affects glucose repression less than the kht1 kht2 double mutation could be explained by the presence (though in reduced levels) of Kht1 in the rag4 mutant, allowing glucose to enter into the cell and to exert repression of some glucose-sensitive genes. It is notable that disruption of the RAG4 gene in the strain that is not glucose sensitive (MW270-7B) led to a strong reduction of RAG1 transcription without completely eliminating low-affinity glucose uptake. In S. cerevisiae the signal for glucose repression appears to be related to the glucose concentration rather than the glucose flux (MEIJER et al. 1998 ), and it seems likely that the intracellular glucose concentration is important (YE et al. 1999 ). Our present findings, together with previous results establishing that glucose repression is not abolished in a phosphoglucose isomerase mutant (WEIRICH et al. 1997 ), are in agreement with the results obtained in S. cerevisiae. Altogether, these data suggest that glucose needs to enter into the cell but does not need to be metabolized to produce the signal for glucose repression.

	FOOTNOTES

¹ Present address: Cancer Research Institute, Slovak Academy of Sciences, Department of Molecular Biology and Biochemistry, Bratislava 842 15, Slovakia.

	ACKNOWLEDGMENTS

We are grateful to Sabira Özcan and Marian Carlson for sharing plasmids. We thank Hiroshi Fukuhara for encouragement and helpful discussions. Svätopluk Betina was a recipient of a Fellowship from the Ministère de l'Education Nationale, de l'Enseignement Supérieur et de la Recherche. This work was supported in part by a grant from the Commission of the European Communities (BIO4-CT96-0003) and in part by a grant from Ministero Università e Ricerca Scientifica e Tecnologica-Università di Parma Cofin 1999.

Manuscript received December 11, 2000; Accepted for publication February 19, 2001.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ALBERTI, A., P. GOFFRINI, I. FERRERO, and T. LODI, 2000  Cloning and characterization of the lactate-specific inducible gene KlCYB2, encoding the cytochrome b2 of Kluyveromyces lacti.. Yeast 16:657-665[Medline].

BILLARD, P., S. MÉNART, J. BLAISONNEAU, M. BOLOTIN-FUKUHARA, and H. FUKUHARA et al., 1996  Glucose uptake in Kluyveromyces lactis: Role of the HGT1 gene in glucose transport. J. Bacteriol 178:5860-5866[Abstract/Free Full Text].

BLAISONNEAU, J., H. FUKUHARA, and M. WÉSOLOWSKI-LOUVEL, 1997  The Kluyveromyces lactis equivalent of caseine kinase I is required for the transcription of the gene encoding the low-affinity glucose permease. Mol. Gen. Genet. 253:469-477[Medline].

BREUNIG, K. D., 1989  Glucose repression of LAC gene expression in yeast is mediated by the transcriptional activator LAC9. Mol. Gen. Genet. 216:422-427[Medline].

BREUNIG, K. D., M. BOLOTIN-FUKUHARA, M. M. BIANCHI, D. BOURGAREL, and C. FALCONE et al., 2000  Regulation of primary carbon metabolism in Kluyveromyces lactis.. Enzyme Microb. Technol. 26:771-780[Medline].

CHEN, X. J., 1996  Low- and high-copy-number shuttle vectors for replication in the budding yeast Kluyveromyces lactis.. Gene 172:131-136[Medline].

CHEN, X. J., M. WÉSOLOWSKI-LOUVEL, and H. FUKUHARA, 1992  Glucose transport in the yeast Kluyveromyces lactis. II. Transcriptional regulation of the glucose transporter gene RAG1. Mol. Gen. Genet. 33:97-105.

COHN, M. and J. MONOD, 1951  Purification et proprieté de la ß-galactosidase d'Escherichia coli.. Biochem. Biophys. Acta 7:153-165[Medline].

COONS, D., M. P. VAGNOLI, and L. F. BISSON, 1997  The C-terminal domain of Snf3p is sufficient to complement the growth defect of snf3 null mutations in Saccharomyces cerevisiae: SNF3 functions in glucose recognition. Yeast 13:9-20[Medline].

DICKSON, R. C., R. M. SHEETZ, and L. R. LACY, 1981  Genetic regulation: yeast mutants constitutive for ß-galactosidase activity have an increased level of ß-galactosidase messenger ribonucleic acid. Mol. Cell. Biol. 1:1048-1056[Abstract/Free Full Text].

DOHMEN, R. J., A. W. M. STRASSER, C. B. HÖNER, and C. P. HOLLENBERG, 1991  An efficient transformation procedure enabling long-term storage of competent cells of various yeast genera. Yeast 7:691-692[Medline].

FERRERO, I., C. ROSSI, M. P. LANDINI, and P. P. PUGLISI, 1978  Role of the mitochondrial protein synthesis in the carbon catabolite repression of the petite-negative yeast K. lactis.. Biochem. Biophys. Res. Commun. 80:340-348[Medline].

GOFFRINI, P., A. A. ALGERI, C. DONNINI, M. WÉSOLOWSKI-LOUVEL, and I. FERRERO, 1989  RAG1 and RAG2: nuclear genes involved in the dependence/independence on mitochondrial respiratory function for the growth on sugars. Yeast 5:99-106[Medline].

GOFFRINI, P., A. FICARELLI, and I. FERRERO, 1995  Hexokinase activity is affected in mutants of Kluyveromyces lactis resistant to glucose repression. Microbiology 141:441-447.

LIANG, H. and R. F. GABER, 1996  A novel signal transduction pathway in Saccharomyces cerevisiae defined by Snf3-regulated expression of HXT6.. Mol. Biol. Cell 7:1953-1966[Abstract].

LODI, T., D. O'CONNOR, P. GOFFRINI, and I. FERRERO, 1994  Isolation and characterization of the KlDLD gene encoding the carbon catabolite repressible mitochondrial enzyme D-lactate ferricytochrome c oxidoreductase in Kluyveromyces lactis.. Mol. Gen. Genet. 244:622-629[Medline].

MARSHALL-CARLSON, L., J. L. CELENZA, B. C. LAURENT, and M. CARLSON, 1990  Mutational analysis of the SNF3 glucose transporter of Saccharomyces cerevisiae.. Mol. Cell. Biol. 10:1105-1115[Abstract/Free Full Text].

MEIJER, M., M. C. J. BOONSTRA, A. J. VERKLEIJ, and C. T. VERRIPS, 1998  Glucose repression in Saccharomyces cerevisiae is related to the glucose concentration rather than the glucose flux. J. Biol. Chem. 273:24102-24107[Abstract/Free Full Text].

NEIGEBORN, L., P. SCHWARTZBERG, R. REID, and M. CARLSON, 1986  Null mutation in the SNF3 gene of Saccharomyces cerevisiae cause a different phenotype than do previously isolated missense mutations. Mol. Cell. Biol. 6:3569-3574[Abstract/Free Full Text].

ÖZCAN, S., J. DOVER, A. G. ROSENWALD, S. WÖLFL, and M. JOHNSTON, 1996  Two glucose transporters in Saccharomyces cerevisiae are glucose sensors that generate a signal for induction of gene expression. Proc. Natl. Acad. Sci. USA 93:12428-12432[Abstract/Free Full Text].

ÖZCAN, S., J. DOVER, and M. JOHNSTON, 1998  Glucose sensing and signaling by two glucose receptors in the yeast Saccharomyces cerevisiae.. EMBO. J. 17:2566-2573[Medline].

PRIOR, C. P., H. MAMESSIER, X. J. FUKUHARA, M. CHEN AND, and M. CHEN ANDWÉSOLOWSKI-LOUVEL, 1993  The hexokinase gene is required for the transcriptional regulation of the glucose transporter gene RAG1 in Kluyveromyces lactis.. Mol. Cell. Biol. 13:3882-3889[Abstract/Free Full Text].

SALMERON, J. M. and S. A. JOHNSTON, 1986  Analysis of the Kluyveromyces lactis positive regulatory gene LAC9 reveals functional homology to, but sequence divergence from, the Saccharomyces cerevisiae GAL4 gene. Nucleic Acids Res. 14:7767-7781[Abstract/Free Full Text].

SCHMIDT, M., C. R. R. MCCARTNEY, X. ZHANG, T. S. TILLMAN, and H. SOLIMEO et al., 1999  Std1 and Mth1 proteins interact with the glucose sensors to control glucose-regulated gene expression in Saccharomyces cerevisiae.. Mol. Cell. Biol. 19:4561-4571[Abstract/Free Full Text].

THIERRY, A., C. FAIRHEAD, and B. DUJON, 1990  The complete sequence of the 8.2 kb segment left of MAT on chromosome III reveals five ORFs, including a gene for a yeast ribokinase. Yeast 6:521-534[Medline].

VAGNOLI, P., D. M. COONS, and L. F. BISSON, 1998  The C-terminal domain of Snf3p mediates glucose-responsive signal transduction in Saccharomyces cerevisiae.. FEMS Microbiol. Lett. 160:31-36[Medline].

WEIRICH, J., P. GOFFRINI, P. KUGER, I. FERRERO, and K. D. BREUNIG, 1997  Influence of mutations in hexose-transporter genes on glucose repression in Kluyveromyces lactis.. Eur. J. Biochem. 249:248-257[Medline].

WÉSOLOWSKI, M., A. A. ALGERI, P. GOFFRINI, and H. FUKUHARA, 1982  Killer DNA plasmids of the yeast Kluyveromyces lactis. I. Mutations affecting the killer phenotype. Curr. Genet. 150:137-140.

WÉSOLOWSKI-LOUVEL, M., P. GOFFRINI, I. FERRERO, and H. FUKUHARA, 1992a  Glucose transport in the yeast Kluyveromyces lactis. I. Properties of an inducible low-affinity glucose transporter gene. Mol. Gen. Genet. 33:89-96.

WÉSOLOWSKI-LOUVEL, M., C. PRIOR, D. BORNECQUE, and H. FUKUHARA, 1992b  Rag^- mutations involved in glucose metabolism in yeast: Isolation and genetic characterization. Yeast 8:711-719.

YE, L., A. L. KRUCKEBERG, J. A. BERDEN, and K. VAN DAM, 1999  Growth and glucose repression are controlled by glucose transport in Saccharomyces cerevisiae cells containing only one glucose transporter. J. Bacteriol. 181:4673-4675[Abstract/Free Full Text].

ZENKE, F., W. ZACHARIAE, A. LUNKES, and K. D. BREUNIG, 1993  Gal80 proteins of Kluyveromyces lactis and Saccharomyces cerevisiae are highly conserved but contribute differently to glucose repression of the galactose regulon. Mol. Cell. Biol. 13:7566-7576[Abstract/Free Full Text].

ZIMMERMANN, F. K. and I. SCHEEL, 1977  Mutants of Saccharomyces cerevisiae resistant to carbon catabolite repression. Mol. Gen. Genet. 154:75-82[Medline].

ZIMMERMANN, F. K., I. KAUFMANN, H. RASENBERGER, and P. HAUSMANN, 1977  Genetics of carbon catabolite repression in Saccharomyces cerevisiae: genes involved in the derepression process. Mol. Gen. Genet. 151:95-103[Medline].

This article has been cited by other articles:

				O. G. Stasyk, M. M. Maidan, O. V. Stasyk, P. Van Dijck, J. M. Thevelein, and A. A. Sibirny Identification of Hexose Transporter-Like Sensor HXS1 and Functional Hexose Transporter HXT1 in the Methylotrophic Yeast Hansenula polymorpha Eukaryot. Cell, April 1, 2008; 7(4): 735 - 746. [Abstract] [Full Text] [PDF]

				S. Rolland, M. Hnatova, M. Lemaire, J. Leal-Sanchez, and M. Wesolowski-Louvel Connection Between the Rag4 Glucose Sensor and the KlRgt1 Repressor in Kluyveromyces lactis Genetics, October 1, 2006; 174(2): 617 - 626. [Abstract] [Full Text] [PDF]

				A. Suleau, P. Gourdon, J. Reitz-Ausseur, and S. Casaregola Transcriptomic Analysis of Extensive Changes in Metabolic Regulation in Kluyveromyces lactis Strains. Eukaryot. Cell, August 1, 2006; 5(8): 1360 - 1370. [Abstract] [Full Text] [PDF]

				F. Abdel-Sater, M. El Bakkoury, A. Urrestarazu, S. Vissers, and B. Andre Amino Acid Signaling in Yeast: Casein Kinase I and the Ssy5 Endoprotease Are Key Determinants of Endoproteolytic Activation of the Membrane-Bound Stp1 Transcription Factor Mol. Cell. Biol., November 15, 2004; 24(22): 9771 - 9785. [Abstract] [Full Text] [PDF]

				M. Lemaire and M. Wesolowski-Louvel Enolase and Glycolytic Flux Play a Role in the Regulation of the Glucose Permease Gene RAG1 of Kluyveromyces lactis Genetics, October 1, 2004; 168(2): 723 - 731. [Abstract] [Full Text] [PDF]

				O. V. Stasyk, O. G. Stasyk, J. Komduur, M. Veenhuis, J. M. Cregg, and A. A. Sibirny A Hexose Transporter Homologue Controls Glucose Repression in the Methylotrophic Yeast Hansenula polymorpha J. Biol. Chem., February 27, 2004; 279(9): 8116 - 8125. [Abstract] [Full Text] [PDF]

				C. Milkowski, S. Krampe, J. Weirich, V. Hasse, E. Boles, and K. D. Breunig Feedback Regulation of Glucose Transporter Gene Transcription in Kluyveromyces lactis by Glucose Uptake J. Bacteriol., September 15, 2001; 183(18): 5223 - 5229. [Abstract] [Full Text] [PDF]